Fiscal Policy Reform in Latin America Max Alier and Benedict Clements International Monetary Fund
This paper was produced for the Consulta de San José 2007 project. The final version of this paper can be found in the book, ‘Latin American Development Priorities - Cost and Benefits’, first edition, edited by Bjørn Lomborg (Cambridge University Press, 2009)
Comments on “Fiscal Policy Reform in Latin America” by Miguel Braun1 Max Alier and Benedict Clements International Monetary Fund September, 2007 Introduction Mr. Braun provides a strong case for his thesis that that the principal fiscal challenge facing Latin America is to build on recent gains in achieving fiscal solvency and reduce the procyclical tendency of fiscal policy. He rightly notes the negative consequences of fiscal insolvency and procyclicality on growth, poverty, and income distribution. He argues that there is no scope for complacency, as recent improvement in the region’s fiscal position owe, at least in part, to unusually favorable external conditions that have contributed to transitory increases in revenues and reduced borrowing costs. At the same time, the author notes that public expenditure is rising during the economic expansion, in line with the region’s legacy of procyclical fiscal policy. Based on a broad review of the literature, the paper argues that insolvency and procyclicality are the result of external shocks, high levels of foreign currency denominated debt, and politico-institutional factors. It proposes the following specific solutions to meet the challenge: issuance of contingent debt instruments, adopting fiscal rules, replacing corporate taxes with personal income taxes, and establishing a policy evaluation agency. It estimates the benefits of these solutions at up to 3 percentage points of GDP per year. We find that we have much common ground with Mr. Braun regarding this assessment of the fiscal challenges facing the region. We especially share his concerns regarding the rapid growth of expenditures. Real noninterest public outlays rose by an average of 7½ percent in 2005-06, with current spending accounting for the bulk of the increase. Expenditure-to-GDP ratios have also risen, reversing the declines realized during the early phases of the recovery. At the same time, we feel that the author has described the challenges too narrowly. In what follows below, we provide a broader, more comprehensive view of the challenges facing fiscal policy. Under our vision, the demands on the Ministry of Finance are even greater than those described by Mr. Braun, as we see fiscal policy as a key instrument for achieving high, sustainable, and equitable growth. The Challenge In our view, Latin America’s main fiscal policy challenge is to address the region’s immense social needs by creating the conditions for sustained and equitable output growth. In this context, consolidating fiscal solvency and reducing procyclicality are necessary elements of any strategy to meet this broader challenge. However, we are less pessimistic than Mr. Braun 1
This “alternative view” paper was prepared for the Copenhagen Consensus for Latin America and the Caribbean—Consulta de San José, Costa Rica, October 20-25, 2007.
AVP_Fiscal_AlierBenedict_2v.doc September 18, 2007 (9:55 PM)
2 regarding the progress made on these fronts in recent years. At the same time, we place greater emphasis that Mr. Braun on the need to improve the efficiency and equity of taxation and spending. Although the solution paper recognizes the relevance of tax policy and the need to strengthen the equity and efficiency of government outlays, we feel that it does not give these issues enough prominence. Fiscal vulnerabilities have been reduced considerably during the recent economic upswing. Some, but not all, of this owes to improvements in the external environment. For example, commodity-based revenues rose by an average of about 4¾ percentage points of GDP between 2002 and 2006, but noncommodity revenues also rose.2 Revenue increases have outpaced the surge in spending, and primary surpluses climbed in 2006 for the fourth year in a row to over 3 percent of GDP. Stronger fiscal balances, in conjunction with solid economic growth, have reduced public debt burdens. On a weighted average basis, public debt to GDP ratios have declined to about 52 percent of GDP, a drop of about 24 percentage points of GDP since 2002. Vulnerabilities have also been reduced by changes in the composition of public debt, with a sharp decline in the share of foreign-currency denominated liabilities. This has been facilitated by the ability of many countries in the region to issue debt in their own currencies, proving that “original sin” can, with sufficient redemption, be forgiven in international capital markets. Although fiscal vulnerabilities have been reduced in recent years, more progress would clearly be desirable. Continued solid performance in containing budget deficits and reducing debt burdens will not, however, be enough to address one of the major shortcomings of government in Latin America: the performance of the state in facilitating high and equitable economic growth. Economic growth in the region still lags behind the developing country average, and social indicators, while improving, suggest a high level of unmet needs and a high degree of inequality. To achieve rapid progress on these fronts, it will be essential that the state provide a more supportive role for economic growth—including through changes in tax and expenditure policy. In what follows below, we discuss some of these challenges:3 Reallocating public spending toward investment and raising the efficiency of these outlays. Public investment in Latin America lags that of other regions of the world (see Table). These outlays have been declining over time, and have fallen as a share of total public spending. The increased role of the private sector in providing infrastructure has helped compensate for this decline. Nevertheless, low rates of public investment have most likely contributed to lags in the region’s infrastructure, as measured, for example, by the quality of its road system.
This description of fiscal developments in the region draws heavily on IMF (2007).
This section draws heavily on Clements, Faircloth, and Verhoeven (2007a, 2007b).
Public Investment in Latin America and Other Regions, 1995-2006 (In percent of GDP) 1995-99
Latin America 1/
Africa Asia Central and Eastern Europe
7.6 8.6 3.6
7.4 8.4 3.6
8.0 8.6 3.6
Sources: Authors' calculations, based on data from national authorities, and the WEO database. 1/ For Latin America, data covers 17 countries.
Inefficiencies in public investment spending may also be contributing to infrastructure shortfalls. The efficiency of these outlays, as measured by the relationship between public investment and improvements in infrastructure indicators, varies markedly across countries (Clements, Faircloth, and Verhoeven, 2007a, 2007b). This suggests that there is scope for improvement, in particular by following the best practices in the region in project selection, evaluation, and monitoring (IMF, 2005). Countries should also implement stable multi-year budgets for public investment and strengthen staff capacity and data for project evaluation (Aldunate, 2007). Improving the effectiveness of the civil service. The typical Latin American country spends about 7 percent of GDP to compensate its public employees (excluding public enterprises), about average for developing regions. There is a wide variety in the quality of the civil service in Latin America, and most countries fall short in attracting sufficiently qualified staff an enforcing good performance (EchebarrĂa and CortĂĄzar, 2006). The quality of government services in the region has stagnated since the late 1990s, and remains below that of other, more dynamic regions. The Quality of the Bureaucracy (ICRG Index, 1990-2005)
Latin America Asia
Eastern Europe & Baltics Subsaharan Africa
2.5 2.0 1.5 1.0 0.5 0.0 1990
Source: The International Country Risk Guide database.
There is ample evidence to suggest that greater value-for-money could be achieved in public wage bills in Latin America. Countries that more handsomely compensate their employees, for example, do not necessarily enjoy more productive bureaucracies (Clements, Faircloth, and Verhoeven, 2007a). Achieving a more productive civil service will require deep-rooted changes in incentive systems and addressing the key weaknesses of civil service systems noted in the comprehensive review of Echebarría and Cortázar (2006)—patronage in hiring and promotions, the absence of performance evaluation, and internal inequities in remuneration. Increasing the efficiency and targeting of social spending. Social spending—defined as expenditure for education, health, pensions, social assistance, and housing—absorbs about half of noninterest government outlays. At 13 percent of GDP, these outlays are lower than in the OECD and Eastern Europe, but higher than Emerging Asia. Education and health indicators in Latin America are broadly in line with the region’s level of development (ECLAC, 2006), but still trail those of high-performing regions. Some of these lags reflect weaknesses in the quality of social spending, including in education. Repetition rates remain high, and performance on international examinations in mathematics, reading, and science indicate that Latin America fares poorly relative to other countries. As in the case of the civil service, improving the efficiency of education spending will require that countries address shortcomings in incentive systems and examine opportunities to improve the composition of spending, which remains heavily weighted to salaries at the expense of other inputs (de Ferranti and others, 2003). Table 7. General Government Social Spending in Latin America and Other Regions, 2004 (In percent of GDP) 1/
Latin America 3/ Emerging Asia 4/ Eastern Europe and Central Asia Middle East and North Africa 5/ Sub-Saharan Africa 6/ OECD
Housing and community amenities
4.2 3.5 4.8 4.2 5.5 6.4
2.6 1.3 4.4 2.0 2.9 6.9
5.4 2.2 12.1 1.8 3.7 17.3
0.9 1.1 1.5 1.1 0.5 0.8
12.7 8.4 22.8 9.1 13.8 32.6
Sources: Government Financial Statistics, IMF; Social Indicators and Statistics database, ECLAC; national authorities; and IMF staff estimates. 1/ 2004 or latest available year. 2/ Number of observations vary by category. Therefore, the total social spending regional averages may not necessarily equal the sum of the regional averages of the spending components. 3/Unweighted averages for seventeen countries. 2004 data for Bolivia, Chile, Colombia, Peru, and Uruguay refer to 2003 data. 2004 data for Argentina are based on the national authorities and IMF staff estimates; and for Honduras, 2002 and 2003 data are based on figures for 2001. For Peru, data for 2002 through 2004 for education and social security refer to 2001 data. For Colombia and El Salvador, data were provided by the national authorities. 4/ The Emerging Asia sample includes Hong Kong SAR, Macao SAR, India, Korea, Singapore, Thailand and Vietnam. Korea and Thailand refer to central government. 5/ Budgetary central government for eight countries. 6/ Central government for seven countries.
Reallocating social spending to programs that most benefit the poor will also be important for forging a more equitable society. The distributive incidence of social spending varies greatly across programs, with primary education and social assistance programs having the most
5 favorable impact, while higher education and social insurance programs tend to benefit middle and upper-income groups. Because of the low share of spending in pro-poor programs—such as social assistance—the majority of social spending benefits accrue to those that are relatively well off. Latin America: Distribution of Benefits from Social Spending to the Top and Bottom Quintiles 1/ Education Primary Secondary Tertiary Health Social security Total social spending
Poorest quintile 20.2 29.0 13.2 1.9 20.6 5.6 15.0
Richest quintile 20.4 7.9 18.3 52.1 17.6 51.2 30.4
Memorandum item: Share of quintiles in primary income
Source: Authors' calculations, based on ECLAC (2006). 1/ Unweighted average. Country coverage varies by category. For total spending, total education, health, and social security spending, the number of countries covered is eight, thirteen, fourteen, and nine, respectively.
The region has demonstrated its capacity to develop efficient and well-targeted social assistance programs, such as Bolsa Familia in Brazil, Oportunidades in Mexico, and Familias in Accion in Colombia. Yet spending on these and other social assistance programs, for all their successes, only reach about 1-1½ of GDP—a modest share of total social spending. Beyond the expansion of these programs, it will also be necessary to take the difficult task of reallocating spending, over time, away from programs that primarily benefit upper-income groups. As suggested in Clements, Faircloth, and Verhoeven (2007a, 2007b), steps in this direction include further reform of public pension systems to place them on an actuarially sound footing; greater reliance on user fees in higher education, combined with more widespread access to scholarships for lower-income students; and improvements in the quality of secondary education to reduce repetition rates and increase access to higher education. Moving towards a progressive and less distortionary tax structure. There are very large differences in tax burdens across the region. They range from the low burdens of countries endowed with nonrenewable resources like Mexico and Venezuela (about 10 percent of GDP), to high levels in countries like Brazil (36 percent of GDP).4 Despite these differences, there are a number of common weaknesses in the tax systems across the region. First, there is heavy reliance on indirect and payroll taxes—representing about 75 percent of the total— which results in a regressive tax system. While in developed countries the tax system helps to reduce inequality, in Latin America it exacerbates it (Gomez Sabaini, 2006). Second, the tax system is a source of significant distortions in resource allocation, and reduces the region’s competitiveness relative to other fast-growing regions such as East Asia. For instance, 4
Data refers to taxes collected by the federal government.
6 numerous tax exemptions (especially for VATs) and various taxes on financial transactions are commonplace. More generally, the structure of the tax system does not appear to be based on efficiency criteria. Impact of the Tax System on Income Distribution Nicaragua Mexico Honduras Argentina Panama El Salvador Chile Ecuador Costa Rica Brazil Japan United States Italy Australia Germany The Netherlands Finland Denmark Belgium Sweden 0
Source: Gomez Sabaini, 2006.
Gini Coefficient After Taxes
Gini Coefficient Before Taxes
The Root Causes of the Regionâ€™s Fiscal Problems5 We are in broad agreement with the author on the root causes behind the regionâ€™s tendency for procyclical policy and periodic bouts of fiscal insolvency. We agree that these problems owe to economic volatility, which is the result of both external volatility and weaknesses in economic policies and strategies that have exacerbated, rather than mitigated, the impact of external shocks. These include a vulnerable structure of debt (with a high share denominated in foreign currency and at short maturities), a low degree of openness to trade, and a history of debt intolerance. While agreeing that these factors have been behind crises in earlier periods, it is important to note that the region has made great progress in reducing these vulnerabilities over the past few years. We also agree that politico-institutional factors have also played a role in making fiscal policy procyclical and, in some instances, unsustainable. Institutional weaknesses in the region have prevented countries from ameliorating the common-pool and principal-agent problems that characterize policymaking. The solution paper proposes the creation of an independent agency to help improve the quality of public spending. We strongly support the idea that governments should provide rigorous and systematic evaluations of the cost-benefit and distributive impact of government 5
This section draws heavily on Alier (2007).
7 projects and programs. However, it is not clear that an independent agency would need to be created, in all cases, to undertake such a task. We believe that the paper would have benefited from paying more attention to other causes underlying the poor quality of fiscal policy. In particular, we believe that the constraints imposed on fiscal management by budget rigidities deserve greater prominence.6 Revenue earmarking and mandatory expenditure requirements together affect a significant share of governmentsâ€™ budgets, ranging from two-thirds of primary revenue in Chile to about 85 percent in Brazil. In our view, addressing the budget rigidities problem is a precondition to meet the challenges identified in the previous section. Budget Rigidity in Selected Latin American Countries (In percent of primary revenue) 90 80 70 60 50 40 30 20 10 0
Brazil Combined rigidity
Ecuador Expenditure rigidity
Some degree of budget rigidity can be justified on theoretical and practical grounds (see Alier, 2007). But on balance, the benefits of rigidities are outweighed by their significantly negative impact on budget management. Constraints to fiscal management introduced by budget rigidities reduce economic efficiency and lead to suboptimal outcomes. The negative consequences of budget rigidities stem from:
Budget rigidities stem from institutional arrangements that limit the budgetary authoritiesâ€™ ability to adjust the composition and size of the budget in the short run. A number of budget components are naturally inflexible like wages, pensions, and debt service. Other inflexibilities are rooted in the constitution, laws, or decrees that earmark revenues, set minimum spending requirements, or link spending to the evolution of certain macroeconomic variables like inflation, growth, or unemployment. Furthermore, some expenditures are rigid for political reasons.
8 Limits to the reallocation of public spending in response to changing needs. These limitations lead to overspending in some sectors at the expense of others. The resulting resource misallocation is likely to worsen over time as social preferences evolve and governments face limitations on shifting resources across sectors to equalize marginal benefits. Latin American governments’ difficulty in addressing emerging infrastructure gaps is a good example of the effect of budget rigidities on the allocation of public resources. A built-in bias towards higher spending. Mandatory spending requirements negatively affect the quality of fiscal adjustment, as the brunt of the spending retrenchment falls on a subset of budget items. Macroeconomic management and political economy reasons can provide reasonable explanations as to why it may be optimal to first cut capital spending in times of rapid fiscal retrenchment. However, after the initial tightening, budget rigidities limit the scope for governments to rebalance expenditure allocations to reflect tighter budget constraints. Budget rigidities also limit the government’s capacity to reallocate resources towards social safety nets at times of economic crisis. Examples of this phenomenon can be found in numerous fiscal adjustment episodes in the region. A built-in bias towards higher taxation. Revenue earmarking also negatively affects the quality of fiscal adjustment by amplifying the contractionary effects of revenue-based fiscal adjustments. Because of requirements to spend at least part of the new revenues for designated purposes, revenue-based fiscal adjustments require sharper increases in taxes. Over the medium term, the additional public spending created during the tightening cycle is difficult to roll back, and governments face difficulties in cutting taxes back to preadjustment levels. Distortions introduced in tax policy choices. Tax policy decisions can be negatively affected by revenue-earmarking considerations. Earmarking arrangements introduce a bias for policymakers to favor revenues that are less earmarked, and governments may try to rebalance the tax burden in this direction. This bias is stronger when taxes are earmarked to expenditures executed by other spending agencies or levels of government, as is the case with revenue sharing agreements in a federation. Examples of how earmarking provisions affected tax policy choices are abundant. In Brazil, the fiscal adjustment in 1999–04 was largely achieved by increasing “social contributions,” the revenues of which are earmarked but not shared with states and municipalities. In Argentina in 2001, the federal government introduced a bank credit and debit tax creditable against VAT and income tax obligations. The latter taxes were shared with the provinces and subject to extensive earmarking, while the newly created tax was not. Eventually, the government agreed to share 30 percent of the tax proceeds with the provinces. Limits on the scope for countercyclical fiscal policy. Extensive revenue earmarking limits the governments’ ability to implement an active countercyclical fiscal policy, that is, being able to save surpluses in good times to offset deficits in bad times. In particular, expenditure requirements indexed to GDP and revenue earmarking limit the government’s ability to commit to increase public saving during cyclical economic expansions.
9 Weak incentives to improve the efficiency of public spending. Incentives to improve the efficiency of public spending are severely affected by the presence of budget rigidities, especially mandated spending provisions. In an ideal budgetary process, spending agencies would compete for resources and the best performing agencies (in terms of achieving the governmentâ€™s goals) would see their share of resources increase. Budget rigidities reduce these incentives as spending agencies can count on a guaranteed share of resources independent of their performance. Wasteful public spending may occur as governments are forced to fulfill mandatory spending requirements or comply with revenue earmarking provisions. Given that the economic costs of extensive budget rigidities outweigh their benefits, the rationale for their existence seems to be rooted in the political process. The literature in public choice economics provides a number of arguments to explain and justify the existence of budget rigidities. 7 These arguments include the use of budget rigidities as a means to: Mitigate agency problems in the provision of public goods. Buchanan (1964) explores the implications of revenue earmarking on tax and expenditure choices under majority voting. Davis and Hayes (1993) and Dhillon and Perroni (2001) show that tax earmarking can play a role in mitigating the agency problem in the provision of public goods. They argue that earmarking can reduce the free riding problem in monitoring spending agencies and consequently increase their accountability to taxpayers. In countries that experience rapid decentralization (e.g. Argentina, Brazil, and Ecuador), mandatory spending provisions on health and education for subnational governments can play a role in ensuring the adequate provision of these goods while accountability mechanisms are still being developed. Set limits on the voracity of Leviathan governments. Tax earmarking has been proposed to limit the governmentâ€™s voracity and as a solution to certain externalities. Brennan and Buchanan (1978) argue that earmarking can be used to discipline a Leviathan government that seeks to maximize the surplus it extracts from taxpayers. Reduce vertical and horizontal externalities in a federation when the tax base and expenditure responsibilities overlap. Wrede (2000) argues that tax earmarking could be used to reduce vertical and horizontal externalities in a federation where the tax base and expenditure responsibilities of different levels of government overlap. In highly decentralized countries, minimum expenditure requirements in health and education for local governments can prevent a free rider problem that would lead to underspending in those sectors. Garner support for tax initiatives. Tax earmarking has also been proposed as a mechanism to garner support for tax initiatives. Bret and Keen (2000) argue that as voters become more cynical about politicians, the set of implementable taxes becomes more limited. In such circumstances, earmarking can serve as a mechanism to secure support for tax initiatives by earmarking the revenues to the provision of a good demanded by segments of the society that 7
Most of the literature on the rationale for budget rigidities focuses on revenue earmarking. However, the conclusions typically also apply to expenditure requirements.
10 would have not supported the tax initiative otherwise. In Brazil, the introduction of the bank debit tax (CPMF) constitutes a good example. Political support for the CPMF was largely achieved by offering to earmark the tax to health spending. The literature on the political economy of public debt also provides insights into the reasons underlying the existence of budget rigidities:8 Fiscal illusion. Fiscal illusion facilitates the emergence of budget rigidities. In such a scenario, different groups in society seek to obtain permanent benefits from public spending without fully internalizing the cost of their actions for themselves or society. At the same time, politicians eager to gain or maintain power (e.g. running for re-election) are willing to yield to pressures even when they are aware of the limited benefits for the economy as a whole. Echeverry et al. (2005) show that budget rigidities can emerge when there is a common pool of budget resources that is allocated through a bargaining process in congress. Strategic considerations. Under political uncertainty, incumbent politicians may use budget rigidities to constrain the behavior of their successors, who may have different preferences. Bret and Keen (2000) show that groups advocating particular policies (e.g. an environmental tax) favor earmarking if the efficiency loss of doing so is outweighed by the value of constraining subsequent policymakers’ decisions. They conclude that politically weak politicians are the most likely to favor revenue earmarking. Another conclusion is that budget rigidities would be more prevalent in more polarized societies. A clear example of this is provided by Chile’s Reserved Copper Law, which requires that 10 percent of the state-owned copper company’s gross revenues be transferred to the armed forces for equipment purchase. Inter- and intra-generational conflicts. These also provide explanations for the emergence of budget rigidities. Intergenerational conflict is largely at the heart of mechanisms that make it difficult for governments to reduce pension benefits. Such inflexibilities range from outright legal obstacles to modifying “acquired rights” to mechanisms that guarantee minimum periodical adjustments of benefits, for example through indexation. On the other hand, intragenerational conflict occurs when different groups in society cannot agree on the burden sharing of macroeconomic adjustment. These groups will try to introduce budget rigidities as a defense mechanism and to minimize the share of the adjustment costs that they bear. It is not surprising, then, that budget rigidities are more common in countries with a history of macroeconomic instability. For example, in Argentina and Brazil, the high degree or revenue earmarking is attributed by some as a legacy of the high inflation years. In sum, budget rigidities are the result of deep political conflict in the budget process. Consequently, a permanent solution to budget inflexibility would require a reform of budget institutions, with a view to reduce incentives and opportunities for new budget rigidities to emerge. 8
References are: Alesina and Perotti (1994); Buchanan and Wagner (1977) on theories of fiscal illusion; Alesina and Tabellini (1990) on using debt as a strategic variable; Alesina and Drazen (1991) on intragenerational conflict; and Veloso (2000) on the tragedy of the commons.
Proposed Solutions and Quantification of Costs and Benefits The quantification of the costs and benefits of fiscal reforms is a difficult task. As a general caveat, we note that some of the solutions proposed by the author may have a once-and-forall impact on the level of GDP, and may not necessarily result in a permanent increase in the rate of growth. It is important to bear this possibility in mind and avoid overestimating the benefits of the proposed solutions. We also have the following comments on the specific solutions: Improving debt management by using contingent instruments. We agree on the estimate of the benefits. However, nonstandard debt instruments, for which there is little demand, can only be placed with a premium. This is a cost that Mr. Braun does not consider explicitly, which can be sizable. For example, in Brazil the government virtually eliminated all its external debt by replacing it with domestic debt. Three costs associated with this process were: (i) domestic debt was placed at a higher interest rate; (ii) it was placed with shorter maturity (i.e. increasing vulnerabilities); and (iii) the government lost the valuation gains associated with the appreciation of the real. Although evaluating the counterfactual is difficult, especially in regards to the latter, the cost of these balance sheet operations could be significant. Consolidating the budget process by adopting fiscal rules. Estimates on the impact of fiscal rules may be too optimistic in the absence of details of the rules to be adopted. We believe that adequately designed macro-fiscal rules could play a useful role in mitigating procyclicality and solvency problems, especially if they focus on controlling expenditure growth during economic upswings. However, these rules should not introduce unnecessary rigidities in the budget process. It is also important to underscore, as Mr. Braun does, that fiscal rules would also need to be accompanied by credible enforcement, and rules cannot be seen as a substitute for political will. Reducing budget rigidities, in our view, could be expected to generate benefits by improving the efficiency of public spending. As indicated in our earlier assessment, there is significant scope to improve the efficiency and equity of outlays for public investment, wages, and social spending. Budget rigidities undoubtedly contribute to some of these inefficiencies, and may be contributing to high levels of spending on current outlays.9 The costs of this solution would be the compensating benefits given to those that benefit from present arrangements.
As has been widely demonstrated in empirical studies, government consumption negatively affects long-term economic growth. See Apergis et al. (2007), Barro (1991), Easterly and Rabelo (1993), Barro and Sala-i-Martin (1995), and, more recently, Edwards (2007)
12 References Aldunate, Eduardo, 2007, “Sistemas de Gestíon de Inversión Pública en América Latina,” presented at the 19th Annual Regional Seminar on Fiscal Policy, Economic Commission for Latin America and the Caribbean, Santiago Chile. Alier, Max, 2007, “Measuring Budget Rigidities in Latin America,” IMF Working Paper, forthcoming (Washington: International Monetary Fund). Apergis, Nicholas, Ioannis Filippidis, and Claire Economidou, 2007, “Financial Deepening and Economic Growth Linkages: A Panel Data Analysis”, Review of World Economics, Volume 143, Number 1, April. Barro, Robert, 1991, Economic Growth in a Cross Section of Countries. Quarterly Journal of Economics 106 (2): 407-444 Barro, Robert, and Xavier Sala-i-Martin, 1995, Economic Growth. McGraw Hill (Boston). Brett, C. and M. Keen, Michael (2000). Political Uncertainty and the Earmarking of Environmental Taxes, Journal of Public Economics 75, pages 315-340. Buchanan, J. M. (1963). The Economics of Earmarked Taxes, Journal of Political Economy, vol. 71, no.5. Clements, Benedict, Christopher Faircloth, and Marijn Verhoeven, 2007a, “Getting Spending Right,” Finance and Development, June 2007, pp. 50-52. ______, 2007b, “Public Expenditure in Latin America: Trends and Key Issues,” CEPAL Review, forthcoming. (Revised version of IMF WP 07/21 under a similar title). Davis, M. L. and K. Hayes (1995) The Demand for Good government, Review of Economics and Statistics 75. De Ferranti, David, Guillermo Perry, Indermit Gill, J. Luis Guasch, William Maloney, Carolina Sanchez-Paramo, and Norbert Schady, 2003, Closing the Gap in Education and Technology (Washington: World Bank). De Ferranti, David, Guillermo Perry, Francisco Ferreira, and Michael Walton, 2004, Inequality in Latin America: Breaking with History? (Washington: World Bank). Dhillon, A. and C. Perroni (2001). Tax Earmarking and Grassroots Accountability, in Economic Letters 72, pages 99-106.
Easterly, William and Sergio Rebelo, 1993, Fiscal Policy and Economic Growth. An Empirical Investigation. Journal of Monetary Economics 39 (1): 417-458 Echebarría, Koldo, and Juan Cortázar, 2006, “Public Administration and Public Employment Reform in Latin America,” in Eduardo Lora, ed., The State of State Reform in Latin America (Washington: Inter-American Development Bank). Economic Commission on Latin America and the Caribbean (ECLAC), 2006, Social Panorama of the Americas (Santiago: United Nations). Edwards, Sebastian, 2007, “Crises and Growth: A Latin American Perspective”, NBER, Working Paper 13019, (Cambridge) International Monetary Fund, 2005, Public Investment and Fiscal Policy—Lessons from the Pilot Country Studies (Washington). Gomez Sabaini, Juan, 2006, “Evolución y Situación Tributaria em América Latina: Una Serie de Temas para la Discusión”, in Oscar Cetrángolo and Juan Gomez Sabaini, Tributación em América Latina: En Busca de una Nueva Agenda de Reformas (Santiago: ECLAC, United Nations) International Monetary Fund, 2007, Regional Economic Outlook: Western Hemisphere (Washington). Lindert, Kathy, Emmanuel Skoufias, and Joseph Schapiro, 2006, Redistributing Income to the Poor and the Rich: Public Transfers in Latin America and the Caribbean (Washington: World Bank). Rodrik, Dani, 2006, “The Social Cost of Foreign Exchange Reserves”, NBER, Working Paper 11952 (Cambridge) Wrede M. (2000). Shared Tax Sources and Public Expenditures, International Tax and Public Finance, pages 7-163.